Duo 2300c - Battery rebuilds?

[Trash80toHP_Mini]

Memory unit's FUBAR ATM, is the duo bat 12V or 24V? If the former, you're gonna need to do quite a bit of hackage. If the latter, maybe not quite so much?

 

[LaPorta]

I can put almost anything together, it simply is a matter of WHAT to put together. This is definitely helpful. You fellas have definitely done some ground-breaking work over the years.

 

[Challenger 1983]

Memory unit's FUBAR ATM, is the duo bat 12V or 24V? If the former, you're gonna need to do quite a bit of hackage. If the latter, maybe not quite so much?

The Duo’s battery is a 12 volt unit

I can put almost anything together, it simply is a matter of WHAT to put together. This is definitely helpful. You fellas have definitely done some ground-breaking work over the years.

Do you think this idea of using that prebuilt pack would work?

 

[LaPorta]

I suppose the question is: is the pack just a bunch of cells, or are there intelligent parts in there that interact with the machine?

 

[Challenger 1983]

I suppose the question is: is the pack just a bunch of cells, or are there intelligent parts in there that interact with the machine?

I have a pack open right next to me. There doesn’t seem to be any intelligent parts other than what looks like two thermistors that connect to their own contacts on the logic board, all of the charge circuitry seems to be on the logic board of the powerbook 

 

[beachycove]

Is there a specific regular care routine for NiMH batteries that extends their life?

There are specific battery reconditioning utilities that can significantly extend the runtime and presumably the life of NiMh cells. Years ago, I managed to revive almost dead batteries (like 30 seconds of runtime) so that they would give an hour or an hour and a half of runtime using these utilities; that battery in your 2300 was one of them. The utilities worked best alongside of freezing the things (physically, as in a freezer or leaving in an unheated space overnight) beforehand, which has an effect on crystals that form in old cells. You just let the thing thaw for a day or so, which (one hopes) also deals with condensation.

Check your 270c for these utilities. If I remember correctly, there is one there from Apple, and one from a third party. They both do much the same thing.

 

[Challenger 1983]

There are specific battery reconditioning utilities that can significantly extend the runtime and presumably the life of NiMh cells. Years ago, I managed to revive almost dead batteries (like 30 seconds of runtime) so that they would give an hour or an hour and a half of runtime using these utilities; that battery in your 2300 was one of them. The utilities worked best alongside of freezing the things (physically, as in a freezer or leaving in an unheated space overnight) beforehand, which has an effect on crystals that form in old cells. You just let the thing thaw for a day or so, which (one hopes) also deals with condensation.

Check your 270c for these utilities. If I remember correctly, there is one there from Apple, and one from a third party. They both do much the same thing.

yes that's right, Battery Amnesia and Apple's own Battery Recondition, I'll try freezing both batteries tonight and see how long of a charge I can get out of them 

 

[Trash80toHP_Mini]

I don't recall those methods, very interesting. I recall a utility that would run batteries fully down to zero state. But that might have been for NiCad packs? Do NiMH cells suffer from memory effect?

There is a reconditioning charger for the Duo bats that work such miracles. Couldn't recall the name/maker, but I have one somewhere hereabouts. Found it in the other thread: VST Conditioning Charger

I suppose the question is: is the pack just a bunch of cells, or are there intelligent parts in there that interact with the machine?

I have a pack open right next to me. There doesn’t seem to be any intelligent parts other than what looks like two thermistors that connect to their own contacts on the logic board, all of the charge circuitry seems to be on the logic board of the powerbook 

There should be four components in the Type III battery if memory serves. According to @bigmessowires and @techknight one of the thingies is a 2-wire EEPROM that tells the Duo which battery type it's charging. As I see it, that'd be Q1 in the diagram in the HowTo PDF link in the thread. May still be out there, dunno though. Malwarebytes tells me it's risky to open, so I'll let someone else take a whack at retrieve it. Links probably on p.4 or maybe on p.5 but haven't got time to track it down for this post before work.

View attachment 20570

The Q1/Q2 circuit is flipped outside the battery case, probably to make the schematic less confusing while identifying the mysterious components.

Main point: you may have to tear open that nice neat pack, snip tabs and wire F1 and F2 in series to be able to use it? What's the charger for that thing? I'm guessing it would enclose the pack for thermal feedback, IIRC the VST Charger enclosed the bat. Does that pack's charger contain the fusing and know exactly what its charging? Maybe there's a two wire EEPROM in it or a switch on the charger for setting parameters?

Has anyone seen a spot welder hack suitable for or adaptable to this project? Gotta be someone crazy enough to have done it on YouBoobTube. :approve: I'm thinking that hacking an electronic flash or building a circuit based on one might do the dirty? ZZAAAPPPP!!!!! 8-o

edit: may have come up with a couple of tricks for soldering connections to the snipped tabs.

 

[sutekh]

I'd really encourage a thorough read through the thread I posted above. Questions about voltage, intelligence, charging: all answered. It's not a 12v pack (charged voltage is 13.5v according to the service manual). Yes, it has some "intelligence" in the form of an ID chip that looks like a small TO-92 package transistor (not a thermistor, although there is one of those too). If the charger (the circuitry for which is in the power adapter) can't read the thermistor or the ID chip, that pack won't charge.

 

[Challenger 1983]

The original apple specs for the type 3 put it at 12 volts 

 

[sutekh]

12v nominal perhaps, but charged voltage per the Service Source is indeed 13.5vdc:



Build a battery that outputs only 12v fully charged and you'll have low-voltage warnings and to look forward to

 

[Challenger 1983]

12v nominal perhaps, but charged voltage per the Service Source is indeed 13.5vdc:

View attachment 39726

Build a battery that outputs only 12v fully charged and you'll have low-voltage warnings and to look forward to

Good point, I may end up following your guide and make a lithium pack 

 

[Trash80toHP_Mini]

12v nominal perhaps, but charged voltage per the Service Source is indeed 13.5vdc:

Makes sense, what's the rating of each cell? That baseline x10 has to be over 12V and maybe a just a bit under 15V given a little inefficiency? Can't believe I missed your thread over the summer, nice work. But that's way beyond the scope of what I can do.

You've inspired me to pull out the toolbox casing the hack from what's now been eight years ago! I'll start hunting down the rest of the parts and maybe have another go at that BTI high capacity NiMH setup vs. OEM NiMH setup testing project when I get a chance. Have you any notions about what the BTI PIC controller board might do to fool the Duo into charging their pack to higher capacity than Apple intended for the OEM packs?

View attachment 20526

 

[LaPorta]

Wow, ask just one question...

You guys are amazing. I must admit a tad bit of jealousy. I'm not any kind og engineering or electrical mind, but I wish I was. I've got to ride your coattails for now...

 

[beachycove]

There is some interesting information (in French) here:

https://forums.macg.co/threads/batterie-pour-powerbook-140-180.1291284/

Scroll down and the Duo battery comes up. Google translate is your friend (assuming that your high school French teacher was not).

 

[Challenger 1983]

There is some interesting information (in French) here:

https://forums.macg.co/threads/batterie-pour-powerbook-140-180.1291284/

Scroll down and the Duo battery comes up. Google translate is your friend (assuming that your high school French teacher was not).

Very interesting, I’ll be buying some cells in the next few months so I can make my 2300c truly portable again 

 

[sutekh]

Good point, I may end up following your guide and make a lithium pack 

Happy to help any way I can. My final post in the linked thread is before I added the boost converter (just a cheap-o pre-built 5A module from Amazon) as a replacement for the 2nd ideal diode, but it's otherwise accurate. Aside from the fixed voltage output caveat outlined above, it works great.

Makes sense, what's the rating of each cell? That baseline x10 has to be over 12V and maybe a just a bit under 15V given a little inefficiency? Can't believe I missed your thread over the summer, nice work. But that's way beyond the scope of what I can do.

You've inspired me to pull out the toolbox casing the hack from what's now been eight years ago! I'll start hunting down the rest of the parts and maybe have another go at that BTI high capacity NiMH setup vs. OEM NiMH setup testing project when I get a chance. Have you any notions about what the BTI PIC controller board might do to fool the Duo into charging their pack to higher capacity than Apple intended for the OEM packs?

TBH, I've never handled a functional NiMH Duo pack. The 3 I own (two type-IIs and one III) have either been cannibalized to create Li-Ion packs or are dead as a nail. It may be as high as 15v, which wouldn't hurt the Duo (remember, the power supply outputs ~24vdc), but I suspect the Service Source data is pretty close at 13.5v. Here's a typical NiMH discharge curve:

That also pretty clearly illustrates why Apple chose 11vdc for their LVC value too. Very little usable power beyond that without risking cell damage.

As far as the speculation above about needing to limit cell mAh ratings to align with the capacities of the day, this is just my 2c, but I can't see any reason that would be necessary. The charge cycle is voltage referenced, as are the power manager's discharge thresholds. If you stuff the highest possible capacity cells into a rebuilt pack (which is precisely what I'd do), it should just take longer to charge at the ~600ma rate and longer to discharge (which is exactly what you want, right?) If the rationale is that with so much capacity, you're rarely likely to fully discharge and would therefore encounter issues with cell memory, well, that's what Amnesia is for. These laptops are in no way power efficient by modern standards. Their early LCD / CCFLs, large transistor CPUs, and spinning rust storage all gobble power. More is more.

 

[Trash80toHP_Mini]

By all reports I've seem, using higher capacity cells doesn't work well at all because the Duo won't charge them fully, hence undercharging memory effects for more bad karma. The higher capacity BTI packs with the PIC controller board replacing the blue thingie in the diagram below (the two-wire EEPROM? ) will charge them fully, No undercharged memory effects for the higher capacity cells and more runtime for the Duo.

View attachment 20581

The PIC controller board would make for a truly intelligent recharging system. Or maybe a bully breaking a less than intelligent charging system to its will?

 

[sutekh]

By all reports I've seem, using higher capacity cells doesn't work well at all because the Duo won't charge them fully, hence undercharging memory effects for more bad karma. The higher capacity BTI packs with the PIC controller board replacing the blue thingie in the diagram below (the two-wire EEPROM? ) will charge them fully, No undercharged memory effects for the higher capacity cells and more runtime for the Duo.

The PIC controller board would make for a truly intelligent recharging system. Or maybe a bully breaking a less than intelligent charging system to its will?

I believe the PIC-based board you're referring to is just BTI's answer to simulating the Apple proprietary ID chip. Look how it's connected. The thermistor and ID chip share a common leg (pin 3): Thermistor between cell D and E and read by the Duo on pin 3 & 4 (left to right), and the ID chip emulation board providing it's data on pins 2 & 3. Looks like they needed some extra juice to run the PIC16C54, which is what your orange wire is doing (that micro's range is  2.5v - 6.25v from Microchip's data sheet, hence only cells H, I, and J are feeding it). Smart on their part to use the incidental ground from pin 3 to energize it vs. tying it to pin 1, which would have manifest as a permanent, parasitic drain. Regardless, it isn't tied to all the cells, or even series positive / negative. It doesn't appear to me to be some special intelligent charging system. It's a bit easier to see what's going in in some of the other diagrams from your thread back in 2013. This one for instance:

View attachment 20645

I can't speak to what experience others have had, only my own somewhat exhaustive bench testing of several Duo chargers, but according to my analysis, the charge cycle is simply voltage referenced and current limiting. When it hits the expected charged voltage for a NiMH pack (around the 13.5vdc previously discussed), it stops providing current. If it were tracking how much power had been supplied relative to the initial pack voltage when connected and then discontinuing the charge sequence after supplying a pre-defined amount of energy (based on the capacity of specific cells used in a type-II or III battery for instance), then my LiIon packs wouldn't work well at all. They're more than 2x the original capacity of an OE NiMH, and they charge all the way up to full capacity no problem.

 

[Challenger 1983]

@sutekh which option did you choose for the charger module?